The New Climate Divide in Agriculture: Who Can Recover, and Who Cannot

The next food-security crisis may not arrive as a single failed harvest, but as a widening gap between farms that can absorb climate shocks and those that cannot. With extreme weather becoming more frequent and destructive, agricultural resilience is emerging as one of the most important tests of climate governance.

A cross-country study published in Sustainability argues that climate change is already weakening Agricultural Production Resilience, or APR, across countries, with the strongest damage coming from extreme floods, followed by extreme heat and drought. The study, authored by Fangyan Bai, Chunyan Li, Qi Ban and Wenya Zhang, synthesizes 76 empirical studies published between 2005 and 2025 to compare how five climate factors affect farming systems across regions, production modes and agricultural structures.

Climate change does not hit all farms equally. Low-latitude regions, rain-fed agriculture and crop-dominated systems face far sharper resilience losses than irrigated, diversified or protected systems, the study finds.

Floods Are Now the Hardest Hit to Farm Resilience

Using network meta-analysis, the authors compare five climate factors: extreme high temperature, extreme drought, extreme flooding, precipitation variability and temperature anomaly. All five significantly reduce APR, but their damage is not equal. Extreme flooding shows the strongest negative effect, with a standardized mean difference of −0.352, followed by extreme heat at −0.327, extreme drought at −0.304, precipitation fluctuation at −0.218 and temperature anomaly at −0.105.

The study's ranking is reinforced by SUCRA values, a statistical measure used to compare the relative intensity of the shocks. Extreme flooding ranks first with a SUCRA value of 92.6, ahead of extreme heat at 78.3 and extreme drought at 69.4. The paper explains that floods damage agricultural production not only by destroying crops, but also by harming facilities, degrading soil fertility and disrupting recovery. Heat and drought remain major threats, but they operate through different channels, including heat stress on crops and restricted water supply.

In many countries, agricultural adaptation still focuses heavily on drought preparedness and water scarcity. Those remain essential, but the evidence suggests that flood resilience must move higher on the agenda, especially where drainage, waterlogging control, post-disaster replanting systems and flood insurance remain weak. Climate adaptation for agriculture is no longer only about surviving dry years; it is also about protecting production systems from too much water arriving too fast.

The Climate Burden Falls Where Farming Is Most Exposed

Low-latitude countries and tropical or subtropical regions suffer much stronger shocks than mid- and high-latitude regions. The pooled effect size for low-latitude countries is −0.452, compared with −0.198 for mid- and high-latitude regions. In low-latitude areas, extreme flooding is the dominant risk; in mid- and high-latitude areas, extreme high temperature is the main threat.

This uneven burden matters because many low-latitude countries are also developing economies with large rural populations, weaker agricultural infrastructure and higher dependence on climate-sensitive farming. The study's evidence base includes 36 countries and regions across Asia, Africa, Europe, the Americas and Oceania, covering both developed economies such as the United States, Germany, France and Japan and developing economies including India, Brazil, South Africa, Thailand and parts of sub-Saharan Africa.

For the Global South, climate shocks are not only environmental events; they are development shocks. When floods, heatwaves or droughts reduce farm resilience, they can affect rural incomes, food prices, fiscal stability, nutrition outcomes and migration pressures. A resilience gap in agriculture can quickly become a governance challenge. This is why climate finance and technology transfer should not be treated as optional development support, but as core instruments for protecting food systems in the regions facing the sharpest climate exposure.

Rain-Fed Farms Are the Weak Link in Food Security

The study also shows that how food is produced strongly shapes how vulnerable it is. Rain-fed agriculture and crop farming suffer far greater climate impacts than irrigated agriculture and livestock systems. The pooled effect size for rain-fed agriculture and crop farming is −0.436, compared with −0.209 for irrigated agriculture and livestock systems. The dominant shock for rain-fed crop systems is extreme flooding, while irrigated and livestock systems are more affected by drought.

Rain-fed farms depend heavily on natural rainfall and have limited control over water availability. When rainfall becomes erratic, when droughts last longer or when floods overwhelm fields, these farms have fewer buffers. Irrigation systems, livestock facilities and protected agriculture can reduce exposure by allowing more control over water, temperature and production conditions. This does not make them immune to climate risks, but it does improve their capacity to absorb shocks.

Agricultural structure also matters. Crop-dominated systems face the strongest climate impacts, with a pooled effect size of −0.468. Balanced crop-livestock systems experience a smaller effect of −0.302, while livestock and protected-agriculture-dominated systems show the weakest effect at −0.185. Within crop farming, the share of rain-fed agriculture is closely linked to vulnerability: high rain-fed systems show a pooled effect size of −0.512, compared with −0.226 for low rain-fed and high-irrigation systems. The authors conclude that irrigation infrastructure acts as a core buffer against drought, precipitation variability, extreme heat and other climate shocks.

Governments do not have to wait for perfect climate models to begin reducing vulnerability. Investments in water-saving irrigation, drainage, rainwater harvesting, flood-control infrastructure, storage, logistics, climate-resilient seeds and diversified farming systems can directly strengthen agricultural resilience. For development banks and private investors, the study points to a large adaptation opportunity: the future of food security will depend heavily on infrastructure and technologies that help farms manage both water scarcity and water excess.

Adaptation Must Be Targeted, Not Generic

The study screens 3268 records and retains 76 valid empirical papers, including 38 Chinese and 38 English studies, from which 405 independent effect sizes are extracted for network meta-analysis. The authors also conduct quality checks, robustness tests and publication bias analysis. Their leave-one-out analysis finds that omitting any single study does not change the core conclusions, while a journal-only sensitivity test preserves the same ranking of climate hazards: extreme flood, extreme heat, extreme drought, precipitation fluctuation and temperature anomaly.

Still, the findings should be read with care. APR is measured differently across the studies included in the meta-analysis, and the authors acknowledge that converting diverse indicators into standardized effect sizes requires assumptions that cannot be fully tested with the available literature. The literature search is also mainly limited to Chinese and English sources, meaning some regional or non-English evidence may be missing. The paper further notes that it does not fully examine economic development, agricultural policy, technological investment or institutional environments, and that its analysis is more static than dynamic over time.

Despite these limitations, the policy direction is compelling. Agricultural climate policy must become more differentiated.

• Low-latitude regions need stronger flood early-warning systems, drainage, post-flood recovery mechanisms and waterlogging-resistant crops.
• Mid- and high-latitude regions need stronger heatwave and drought preparedness, water-saving agriculture and smart irrigation.
• Countries with high reliance on rain-fed farming need accelerated investment in water infrastructure, while crop-dominated economies need diversification strategies that reduce concentrated climate risk.

The study also calls for stronger transnational cooperation on agricultural climate monitoring, data sharing, early warning, grain reserves, trade adjustment, technology transfer and green finance support for low-latitude developing regions.

The world needs anticipatory adaptation: building agricultural systems that can absorb shocks before they become food crises. Floods, heat and drought will not affect all regions equally, and resilience will not emerge automatically. It will depend on choices made now about water systems, farm structure, technology, insurance, infrastructure and international cooperation.